Chatbot Sam
I think a common misconception is that citizens (customers) want to have an ‘experience’ when dealing with government agencies. As a customer of a range of agencies what I really want is a pain-free experience that let’s me do what I need to get done in the most expedient manner possible. Get in, get it sorted and get out - as quickly as possible, no experience thanks!
Part of a creating what I refer to as a ‘low friction experience’ is understanding how your customers already utilise online/digital services and slipstreaming in as much as possible with that. The idea of meeting the customer where they are at, rather than setting up another whole new channel for them to engage with.
We know the majority of our customers are part of one or multiple social network platforms. We also know that most of them would list ‘talking to another person on the phone’ as one of their least preferred ways of getting anything done. Based on that it makes solid sense for us to develop a chatbot interface we can deploy on the various social network platforms.
Enter Sam - Sam is our self-service chatbot which provides some very basic functionality but due to the magic of cloud only took hours (not days or months) to set up. Sam is currently in early beta but can help you look up your student number, reset your password and provide you with information on the exam and results publication timetable.
Sam is built on a combination of Amazon’s Lex and Lambda services. Lex provides the chatbot front end whilst Lambda does the behind the scenes functions like interacting with the database hosted on AWS to look up student numbers and other information. We have currently deployed Sam on the Facebook platform meaning it integrates seamlessly with Facebook pages and works on both desktop and mobile devices.
Sam is a pretty basic chatbot - it works based on key terms and rules - however it provides some really good self-service functionality given that it took just a few hours to set up.
The next iteration of Sam will combine Natural Language Understanding (NLU), Automatic Speech Recognition (ASR) and Machine Learning to provide a much richer set of functionality as well as the ability to learn from the conversations it has. We also intend to use Amazon Poly to support Speech Input.
For me Sam re-enforces the power of the Cloud. Five years ago the investment required - and lead times to implement - would have probably seen the idea dead in the water. Today we can deliver a whole new friction-less channel to our customers in a matter of hours.
We are shifting away from running technology to delivering digital services and that’s the exciting opportunity for organisations who are ready and willing to embrace the Cloud.
CEOs need to be challenging their CIOs and IT departments to deliver these same sorts of customer-focused solutions; I’m convinced that will move the cloud adoption conversation forward in most organisations that may be ‘stuck’ in the enterprise IT paradigm.